Bright and stable dye-doped fluorescent polystyrene microspheres: quantitative lateral flow immunoassay for detecting serum amyloid A

Abstract

Difluoroboron is a fundamental molecular building block that enables the development of numerous highly bright fluorophores for biosensing and imaging. However, its poor stability and limited brightness in aqueous solutions are long-standing and unresolved issues in developing difluoroboron-based fluorophores, which significantly limit their widespread biosensing applicability. Herein, we report a generalizable strategy for synthesising dye-doped fluorescent polystyrene microspheres as a local hydrophobic microenvironment to “protect” difluoroboron curcuminoids from water and reactive oxygen species. This strategy demonstrates a breakthrough in the chemical stability, photo-stability and brightness of difluoroboron-based fluorophores in aqueous solutions, making it suitable for point-of-care testing. The results of lateral flow immunoassay using the dye-doped fluorescent polystyrene microspheres highlight a wide linear range, low limit of detection, good intra-assay and inter-assay coefficient of variations; thus, the reported microspheres can be considered an efficient tool for serum amyloid A detection in human serum samples.

---

Introduction

Organic luminescent compounds are of paramount importance in various fields such as bioassays,^1–4 bioimaging,^5–7 and point-of-care testing (POCT).^8,9 Difluoroboron curcuminoid derivatives,^10–13 characterised by their ultra-high brightness, are widely used in biosensing and bioimaging.^14–17 Unfortunately, this dye is extremely sensitive to high polarity microenvironments and oxidation stress, leading to inherent disadvantages in aqueous solutions, including insufficient brightness and poor chemical/photo-stability.^18,19 In this regard, a local hydrophobic microenvironment must be developed to “protect” difluoroboron-based fluorophores from water and reactive oxygen species (ROS). We envision this strategy could achieve synergistic enhancement in the stability and brightness of difluoroboron curcuminoids, thereby improving their practicability in biosensing and bioanalysis.

Herein, we describe a dye-doped polystyrene (PS) microsphere strategy relying on the local hydrophobic microenvironment to “protect” difluoroboron curcuminoids to enhance their stability and brightness. By harnessing this strategy, difluoroboron curcuminoid (BFOH) dyes are encapsulated within PS microspheres, thereby bestowing the following extraordinary features: (i) excellent long-term storage stability (with good chemical stability) for at least 120 days, (ii) ultrahigh photostability against reactive oxygen species attack and (iii) enhanced fluorescence quantum yield (86.4%) compared with free dyes (54.1%) in water (Fig. 1a). Based on these significant stability and brightness improvements, BFOH@PS was successfully applied in lateral flow immunoassay (LFIA) for serum amyloid A (SAA) detection in clinical serum samples (Fig. 1b), showing a wide linear range, low limit of detection (LOD), and good intra-/inter-assay coefficient of variations. In this regard, we provide an appealing paradigm that allows us to design fluorescent PS microspheres for high-fidelity biosensing and POCT.

Fig. 1 Doping difluoroboron curcuminoid dye (BFOH) into polystyrene microspheres: a breakthrough in stability and bright fluorescence for lateral flow immunoassay (LFIA). (a) Schematic representation of the preparation and the local hydrophobic microenvironment of BFOH@PS. (b) SAA antibody (SAA mAb2) coupled with BFOH@PS and the application in LFIA.

Results and discussion

Preparation of uniform difluoroboron curcuminoid-doped polystyrene microspheres with high brightness

Carboxylated polystyrene (PS) microspheres were synthesized in a single step via soap-free polymerization.^20,21 Subsequently, we utilized the swelling method^22,23 to dope difluoroboron curcuminoid dye (BFOH) into PS microspheres: using a swelling agent (THF), the PS chains were swelled and allowed BFOH dyes to penetrate the interior of the microspheres. Then, upon THF solvent evaporation, PS chains contracted and capsulated the BFOH dyes within the PS microspheres (the dye loading in the microspheres was 316 mg g⁻¹) (Fig. S15).^24,25 As shown in Fig. 2a, b and Fig. S13, the PS microspheres and BFOH@PS exhibited uniform and smooth spherical morphologies, with average diameters of 193 ± 2 and 210 ± 2 nm, respectively (Fig. 2c). Furthermore, the larger particle size of BFOH@PS might be attributed to the loading of BFOH dyes (Fig. S14). As shown in Fourier transform infrared (FTIR) spectra (Fig. S7), after doping with BFOH, BFOH@PS exhibited the characteristic peaks of B–F (1147 cm⁻¹) and B–O (1147 cm⁻¹), attributable to BFOH, along with the hydroxyl characteristic peak associated with –OH (3424 cm⁻¹), and this evidence confirmed that BFOH was successfully doped within the PS microspheres. Dynamic laser scattering (DLS) showed that the polydispersity index (PDI) of BFOH@PS was 0.038, indicating its uniform microsphere morphology. More importantly, there was almost no aggregation or disruption of microspheres pre- and post-doping. To further confirm that BFOH dyes were stably capsulated within the PS microspheres (instead of being loosely adsorbed on the surface of the PS microspheres), we exposed BFOH@PS to a water–dichloromethane (DCM) system since PS microspheres are well miscible in water and BFOH is highly soluble in DCM. As shown in Fig. 2d, the color and fluorescence (from BFOH) only appeared in the water phase rather than in DCM, highlighting that nearly no free BFOH dyes were leaked from the PS microspheres. All these results demonstrated the uniform structure of BFOH@PS with significantly stable dye-encapsulation.

Fig. 2 Dye-doped fluorescent microspheres for brightness and stability enhancement. (a and b) TEM images of PS and BFOH@PS. (c) Hydrodynamic size of PS (193 ± 2 nm) and BFOH@PS (210 ± 2 nm). (d) Photographs of BFOH and BFOH@PS in a dichloromethane–water system under room light (left) and ultraviolet light (right). (e) Fluorescence spectra (λ_ex = 480 nm) of BFOH and BFOH@PS in water. The fluorescence quantum yield (Φ) of BFOH and BFOH@PS was measured in PBS (10 mM, pH 7.4). (f) Fluorescence spectra (λ_ex = 480 nm). (g) Zeta potential of PS, BFOH@PS, and BFOH@PS–SAA mAb2. (h) Hydrodynamic size (solid line) and PDI (dashed line) of PS and BFOH@PS after 120 days of storage in water. (i) Continuous monitoring of the fluorescence intensity (at 540 nm) of BFOH@PS for 90 days (λ_ex = 480 nm) in water. (j) The pH-dependent fluorescence intensity of BFOH (at 590 nm) and BFOH@PS (at 540 nm), λ_ex = 480 nm, in water. (k) Fluorescence intensity of BFOH (at 590 nm) and BFOH@PS (at 540 nm) with and without H₂O₂ (72 mM), λ_ex = 480 nm. Data with error bars are expressed as mean ± s.d., n = 3. (l) Time-dependent fluorescence intensity of BFOH (at 590 nm) and BFOH@PS (at 540 nm) under illumination (in water). Sample tubes were exposed to a tungsten–halogen lamp (400–600 nm spectral range, 100 mW cm⁻² power density, tubes were at a distance of 20 cm from the lamp).

To confirm the local hydrophobic microenvironment of BFOH@PS, we carefully compared the fluorescence spectra difference between BFOH and BFOH@PS in water. As shown in Fig. 2e, BFOH@PS displayed a remarkable blue-shifted emission wavelength of around 50 nm compared with free BFOH. It could be interpreted that BFOH entered the hydrophobic microenvironment (long-chain structure of polystyrene), resulting in a decrease in the conformational freedom and rotational restriction of the BFOH fluorophore.^26,27 Moreover, this rotational restriction indeed led to a huge enhancement in the fluorescence quantum yield (Fig. S8) of BFOH@PS (86.4%), much higher than free BFOH dyes (54.1%). Furthermore, the fluorescence spectra of BFOH@PS was more consistent with free BFOH in organic solvents with low polarity (such as THF and DCM), rather than high-polarity ethyl alcohol or DMSO (Fig. 2f). Obviously, these characteristics strongly confirmed the local hydrophobic microenvironment of BFOH@PS, which effectively prevents the collision between water and BFOH dyes, thereby greatly improving the brightness (in aqueous solution) of difluoroboron curcuminoid dye.

Dye-doped PS microspheres improve storage as well as chemical and photo-stability

Subsequently, we explored the PS microsphere encapsulation strategy for enhancing the stability of difluoroboron curcuminoid dyes, particularly focusing on long-term storage, chemical stability, and photo-stability. Long-term storage results (Fig. 2h) indicated that the particle size and PDI (at 120 days) remained unchanged compared with those of the freshly prepared samples. This might be attributed to the presence of negatively charged carboxyl groups on the surface of PS and BFOH@PS (with zeta potentials of −26.9 and −25.97 mV, respectively) (Fig. 2g), in which electrostatic repulsion could inhibit undesirable aggregation and maintain satisfactory dispersity. Furthermore, the fluorescence intensity of BFOH@PS fluctuated slightly over 90 days (Fig. 2i), further confirming the excellent long-term storage stability. To further verify the chemical stability enhancement with dye-doped fluorescent microspheres, we systematically evaluated the BFOH@PS and free BFOH dyes’ stability under aqueous solutions at a wide pH range and using hydrogen peroxide (H₂O₂). Specifically, the PS microsphere encapsulation strategy effectively preserves the stability of BFOH's fluorescence signal and enhances its resistance to pH induced interference. The fluorescence intensity of free dyes was sharply decreased to 14% from pH 1 to 13. In contrast, BFOH@PS maintained high brightness (with 83%) at pH 13 (Fig. 2j). Moreover, the stability of BFOH@PS was assessed in the presence of high concentrations of H₂O₂. After ten-minute incubation, BFOH@PS retained 96% of its initial fluorescence intensity, whereas BFOH retained only 61% (Fig. 2k). Clearly, BFOH@PS exhibited much better chemical stability than that of free BFOH dyes. This is mainly due to the fact that the difluoroboron curcuminoid dyes were encapsulated within the PS microspheres, which ensured well stability in harsh chemical environments. After confirming the good storage and chemical stability of BFOH@PS, we studied its photostability.^28–30 We irradiated BFOH@PS and free BFOH dyes (dispersed in aqueous solutions) using a W-halogen lamp (white light, 350 W) for 60 min. As shown in Fig. 2l, free BFOH dyes demonstrated rapid fluorescence reduction, indicating significant time-dependent photofading. Especially, only 35% of the fluorescence intensity of free dyes was retained after 60 min irradiation, mainly owing to photooxidative degradation. In contrast, 81% of the fluorescence intensity of BFOH@PS was retained after 60 min irradiation. This significantly enhanced photostability of BFOH@PS further demonstrated that the local hydrophobic microenvironment could well “protect” the conjugate structure of BFOH from reactive oxygen species attack.

Preparation of BFOH@PS–SAA mAb2 for LFIA detection

As an acute-phase protein, serum amyloid A (SAA) is highly correlated with various inflammatory diseases, including coronavirus disease (COVID-19).^31–33 Following the onset of the acute-phase response, the blood concentration of SAA rapidly increases by 1000-fold within 24 hours. Currently, SAA detection primarily relies on immunofluorescence, immunoturbidimetry, and chemiluminescent assays. However, these methods inevitably require expensive equipment and reagents, which are unsuitable for POCT in resource-limited settings.^34–36 Therefore, there is an urgent need for an affordable, convenient, rapid, and sensitive diagnostic tool for SAA. In this case, we envision that LFIA³⁷ (with the double-antibody sandwich method,^38,39Fig. 3a–c) could make a breakthrough in SAA detection: (i) when a sample with the SAA antigen was added, SAA immunofluorescent microspheres (BFOH@PS–SAA mAb2) bound to it, forming a complex that was captured on the T line, which emitted a green fluorescent band under light excitation (the buffer used in running LFIA is shown in Table S1). (ii) As the concentration of the SAA antigen increases, the fluorescence intensity of the T line gradually increases and then levels off (Fig. S12). (iii) In the absence of the SAA antigen in the sample, all fluorescent probes are captured at the C line. More importantly, the LFIA test strip (Fig. 3a) could be conveniently performed using a commercial portable LFIA strip reader (reader working mechanism, Fig. 3b).

Fig. 3 Illustration of the BFOH@PS–SAA mAb2 for SAA quantitative detection via the lateral flow immunoassay (LFIA) method. (a) Lateral flow immunoassay test strip components. (b) Schematic of the portable LFIA strip reader, including an LED lamp excited at 480 nm, a complementary metal oxide semiconductor (CMOS) camera, and a 540 nm filter. (c) Schematic of the developed LFIA test strip for the detection of the SAA antigen in human serums.

BFOH@PS exhibits remarkable storage, chemical stability, and photo-stability, and this inspired us to apply it in LFIA for SAA detection. We made full use of the EDC–NHS method^40,41 to activate the carboxyl groups on the BFOH@PS surface and conjugated microspheres with the SAA mAb2 antibody to obtain BFOH@PS–SAA mAb2 immunofluorescent microspheres (the microspheres were stored in the BB buffer, Table S2). The diameter of BFOH@PS–SAA mAb2 was 220 ± 5 nm with a PDI of 0.055 (Fig. S9), and the microsphere's zeta potential changed from −25.97 mV to −33.3 mV (Fig. 2g). These results indicated that SAA mAb2 was successfully modified on the surface of BFOH@PS without aggregation. Notably, we optimized the activation pH, coupling pH, EDC and NHS concentration, and SAA mAb2 content to achieve optimal LFIA strip sensitivity. Based on the T line's fluorescence intensity and a positive sample SAA antigen concentration of 50 mg L⁻¹, the optimal experimental conditions were determined to be pH 6.0, 500 mM MES buffer for activation (Fig. S10a); pH 7.0, 500 mM MES buffer for coupling (Fig. S10b), EDC and NHS concentration of 10 mg mL⁻¹ (Fig. S10c), and 15 μg of SAA mAb2 (Fig. S10d). The optimal immunoreaction time on the LFIA strip was around 10 minutes (Fig. S11). Optimization of these conditions would guarantee the detection accuracy of the test strip.

Highly bright and stable BFOH@PS–SAA mAb2 for quantitative detection of SAA in clinical samples

As mentioned above, LFIA could be conveniently performed using a commercial and portable LFIA strip reader (Fig. 4a and b). As shown in Fig. 4b–d, the LFIA strip for SAA detection demonstrated a strong correlation within the concentration range of 0.7–99.3 mg L⁻¹, whether based on the fluorescence intensity of the T line (Fig. 4c) or the ratiometric values of T/(T + C) (Fig. 4d). These two methods displayed R² values of 0.989 and 0.986, highlighting the quantitative detection performance of BFOH@PS–SAA mAb2. The LOD for SAA detection was calculated as 0.62 mg L⁻¹, which was 16 times less than the clinical threshold (10 mg L⁻¹). More importantly, the intra-assay coefficient of variation (CV) was below 10%, and the inter-assay CV was under 15% (Fig. 4e), indicating the high precision of LFIA. Thus, in terms of the detection range, LOD, and intra-/inter-assay CV, the high brightness and stability of BFOH@PS make it well-suited to meet the requirement of SAA quantitative detection in clinical samples.

Fig. 4 Quantitative detection of SAA in clinical samples using BFOH@PS–SAA mAb2. (a) Images of the LFIA strip (using BFOH@PS–SAA mAb2) under 365 nm ultraviolet light using a portable LFIA strip reader. (b) The height of the fluorescence peak of the LFIA strip was measured using the reader (with different SAA concentrations). (c) The standard curve at log (fluorescent intensity of T line) versus log(SAA concentration). The limit of detection (LOD = mean_blank + 3 × SD) for SAA was based on 10 times the standard deviation of the blank. (d) Standard curve at log(ratiometric values of T/(T + C)) versus log(SAA concentration). (e) Intra-assay and inter-assay precision of the LFIA strip in SAA detection (using BFOH@PS–SAA mAb2). (f) Specificity of BFOH@PS–SAA mAb2 LFIA strip towards SAA (10 mg L⁻¹) using potential interference inflammatory biomarkers, including PCT (300 μg L⁻¹), HBP (3 mg L⁻¹), CRP (10 mg L⁻¹), and IL-6 (10 μg L⁻¹). Data with error bars are expressed as mean ± s.d., n = 3. (g) Immunoturbidimetry method (data from Shanghai Shibei Hospital) and our developed method to detect SAA concentration in clinical serum samples. (h) Correlation coefficient between the immunoturbidimetry method and our developed method.

Subsequently, the specificity of the LFIA strip was evaluated in the presence of four interfering substances (PCT, HBP, CRP, and IL-6) alongside the SAA antigen (Fig. 4f). Results indicated that these interfering substances produced almost no fluorescence, suggesting the negligible impact of these substances. Furthermore, we compared the detection results of an immunoturbidimetry method (data from Shanghai Shibei Hospital, Table S3) and our developed method. A total of 25 human serum samples were analyzed using our developed BFOH@PS–SAA mAb2 LFIA strip, and the results showed close agreement with those obtained by the hospital's immunoturbidimetric method, yielding a high correlation coefficient (R² = 0.991) between the two methods (Fig. 4g and h). All these results demonstrated that the dye-doped fluorescent polystyrene microspheres could well enhance the dye's stability and brightness, significantly improving its applicability and reliability for quantitative SAA detection in clinical serum samples.

Conclusion

In summary, we aimed to dope dyes into PS microspheres to improve the stability and brightness of difluoroboron curcuminoid–derived fluorophores in the aqueous phase. In this strategy, polystyrene long chains can well encapsulate the dye, leading to excellent long-term storage stability (with good chemical stability) for at least 120 days. Furthermore, this microenvironment “protects” the conjugate structure of dyes from reactive oxygen species attack, enhancing the photostability of difluoroboron curcuminoid derivatives. More importantly, the local hydrophobic microenvironment could efficiently decrease the conformational freedom and rotational restriction of BFOH fluorophores, thereby remarkably enhancing the fluorescence quantum yield (86.4%) compared with that of free dyes (54.1%) in water. This significant stability (including storage, chemical stability, and photo-stability) and brightness improvements make it suitable for POCT. Using a portable LFIA strip reader, dye-doped PS microspheres were successfully utilized for accurate SAA detection in clinical serum samples. LFIA results showed a wide linear range (0.7–99.3 mg L⁻¹), low LOD value (16 times lower than the clinical diagnostic value) as well as good intra-assay (CV < 10%) and inter-assay (CV < 15%) CV. We anticipate that this generalizable strategy could also achieve high performance with other molecular dyes and in different biosensing modes, including cell imaging and in vivo tracking.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Experimental procedures and all the relevant data are available within the paper and from the authors.

General methods and supporting data (synthesis and characterizations, ¹H NMR spectrum, ¹³C NMR spectrum, and HRMS spectrum), supporting figures, and supporting references are provided in the SI (PDF). See DOI: https://doi.org/10.1039/d5tb00552c

Acknowledgements

This work was supported by NSFC, China (32394001, 22225805, 32121005, and 22378122); Shanghai Science and Technology Innovation Action Plan (No. 23J21901600); Shanghai Frontier Science Research Base of Optogenetic Techniques for Cell Metabolism (Shanghai Municipal Education Commission, grant 2021 Sci & Tech 03-28); Shanghai Pujiang Program (22PJ1411800); and Science and Technology Commission of Shanghai Municipality (24DX1400200).

Notes and references

1. B. Shen, K. H. Jung, S. Ye, C. A. Hoelzel, C. H. Wolstenholme, H. Huang, Y. Liu and X. Zhang, Aggregate, 2022, 4, 301.
2. F. Wang, Y. Zhong, O. Bruns, Y. Liang and H. Dai, Nat. Photonics, 2024, 18, 535–547.
3. C. Yan, L. Shi, Z. Guo and W. H. Zhu, Chin. Chem. Lett., 2019, 30, 1849–1855.
4. D. Ma, H. Bian, M. Gu, L. Wang, X. Chen and X. Peng, Coord. Chem. Rev., 2024, 505, 215677.
5. J. An, K. Kim, H. J. Lim, H. Y. Kim, J. Shin, I. Park, I. Cho, H. Y. Kim, S. Kim, C. McLean, K. Y. Choi, Y. Kim, K. H. Lee and J. S. Kim, Nat. Commun., 2024, 15, 1004.
6. L. Li, F. Xiang, L. Yao, C. Zhang, X. Jia, A. Chen and Y. Liu, Bioorg. Med. Chem., 2023, 92, 117410.
7. L. Yu, S. A. Abbas Abedi, J. Lee, Y. Xu, S. Son, W. Chi, M. Li, X. Liu, J. H. Park and J. S. Kim, Angew. Chem., Int. Ed., 2023, 62, 202307797.
8. M. Supianto, D. K. Yoo, H. Hwang, H. B. Oh, S. H. Jhung and H. J. Lee, ACS Sens., 2024, 9, 1321–1330.
9. T. Wu, Q. Zhou, G. Chen, S. Huang and G. Ouyang, Analysis Sensing, 2024, 4, 3.
10. C. A. DeRosa, S. A. Seaman, A. S. Mathew, C. M. Gorick, Z. Fan, J. N. Demas, S. M. Peirce and C. L. Fraser, ACS Sens., 2016, 1, 1366–1373.
11. Z. Zeng, J. Ouyang, L. Sun, C. Zeng, F. Zeng and S. Wu, Anal. Chem., 2020, 92, 9257–9264.
12. Y. Zhao, H. S. Kim, X. Zou, L. Huang, X. Liang, Z. Li, J. S. Kim and W. Lin, J. Am. Chem. Soc., 2022, 144, 20854–20865.
13. M. Zhu, T. Wang, Y. Zhang, J. Nong and Z. Guo, Sens. Actuators, B, 2023, 391, 134039.
14. K. Kamada, T. Namikawa, S. Senatore, C. Matthews, P. F. Lenne, O. Maury, C. Andraud, M. Ponce-Vargas, B. Le Guennic, D. Jacquemin, P. Agbo, D. D. An, S. S. Gauny, X. Liu, R. J. Abergel, F. Fages and A. D'Aléo, Chem. – Eur. J., 2016, 22, 5219–5232.
15. J. Miao, Y. Huo, G. Yao, Y. Feng, J. Weng, W. Zhao and W. Guo, Angew. Chem., Int. Ed., 2022, 61, 202201815.
16. K. X. Teng, D. Zhang, B. K. Liu, Z. F. Liu, L. Y. Niu and Q. Z. Yang, Angew. Chem., Int. Ed., 2024, 63, 202318783.
17. J. Zhu, N. K. Tan, K. Kikuchi, A. Kaur and E. J. New, Analysis Sensing, 2023, 4, 202300049.
18. D. Insuasty, L. Cabrera, A. Ortiz, B. Insuasty, J. Quiroga and R. Abonia, Spectrochim. Acta, Part A, 2020, 230, 118065.
19. H. Weiss, J. Reichel, H. Gorls, K. R. A. Schneider, M. Micheel, M. Prohl, M. Gottschaldt, B. Dietzek and W. Weigand, Beilstein J. Org. Chem., 2017, 13, 2264–2272.
20. R. Chen, C. Ren, M. Liu, X. Ge, M. Qu, X. Zhou, M. Liang, Y. Liu and F. Li, ACS Nano, 2021, 15, 8996–9004.
21. X. F. Hu, P. F. Zhang, D. Wang, J. Jiang, X. M. Chen, Y. Liu, Z. W. Zhang, B. Z. Tang and P. W. Li, Biosens. Bioelectron., 2021, 182, 113188.
22. H. Xu, L. Yuan, Q. Shi, Y. Tian and F. Hu, Nano Lett., 2024, 24, 1367–1375.
23. M. Feng, J. Chen, J. N. Xun, R. X. Dai, W. Zhao, H. Z. Lu, J. Xu, L. Chen, G. D. Sui and X. J. Cheng, ACS Sens., 2020, 5, 2331–2337.
24. A. B. Descalzo, H. J. Xu, Z. L. Xue, K. Hoffmann, Z. Shen, M. G. Weller, X. Z. You and K. Rurack, Org. Lett., 2008, 10, 1581–1584.
25. Q. Zhang, Y. Han, W.-C. Wang, L. Zhang and J. Chang, Eur. Polym. J., 2009, 45, 550–556.
26. S. Ye, C. H. Hsiung, Y. Tang and X. Zhang, Acc. Chem. Res., 2022, 55, 381–390.
27. L. L. Yang, H. Wang, J. Zhang, B. Wu, Q. Li, J. Y. Chen, A. L. Tang, J. W. Y. Lam, Z. Zhao, S. Yang and B. Z. Tang, Nat. Commun., 2024, 15, 999.
28. M. Chang, C. Yan, L. Shi, D. Li, W. Fu and Z. Guo, Chin. Chem. Lett., 2022, 33, 762–766.
29. L. Jiao, Y. Liu, X. Zhang, G. Hong, J. Zheng, J. Cui, X. Peng and F. Song, ACS Cent. Sci., 2020, 6, 747–759.
30. C. Yan, H. Xu, M. Wu, Z. Zhao, W. Zhao, J. Tang and Z. Guo, Sci. China: Chem., 2021, 64, 2045–2052.
31. A. A. Abbas, A. Alghamdi, S. Mezghani, M. Ben Ayed, A. M. Alamori, G. A. Alghamdi, W. Bajhmom, H. Wajeeh, S. S. Almutairi, W. M. Radwan and S. R. Bonam, J. Immunol. Res., 2022, 6336556.
32. N. Khan, C. J. Cromer, M. Campa and E. F. Patz, Cancer, 2004, 101, 379–384.
33. G. H. Sack, Mol. Med., 2018, 24, 24–46.
34. A. Sena-Torralba, R. Alvarez-Diduk, C. Parolo, A. Piper and A. Merkoci, Chem. Rev., 2022, 122, 14881–14910.
35. J. Wang, C. Jiang, J. Jin, L. Huang, W. Yu, B. Su and J. Hu, Angew. Chem., Int. Ed., 2021, 60, 13042–13049.
36. M. Xiao, F. Tian, X. Liu, Q. Zhou, J. Pan, Z. Luo, M. Yang and C. Yi, Adv. Sci., 2022, 9, 2105904.
37. F. Di Nardo, M. Chiarello, S. Cavalera, C. Baggiani and L. Anfossi, Sensors, 2021, 21(15), 5185.
38. C. Parolo, A. Sena-Torralba, J. F. Bergua, E. Calucho, C. Fuentes-Chust, L. Hu, L. Rivas, R. Álvarez-Diduk, E. P. Nguyen, S. Cinti, D. Quesada-González and A. Merkoçi, Nat. Protoc., 2020, 15, 3788–3816.
39. G. Zhang, Z. Huang, L. Hu, Y. Wang, S. Deng, D. Liu, J. Peng and W. Lai, ACS Nano, 2023, 17, 23723–23731.
40. W. Wan, Y. Huang, Q. Xia, Y. Bai, Y. Chen, W. Jin, M. Wang, D. Shen, H. Lyu, Y. Tang, X. Dong, Z. Gao, Q. Zhao, L. Zhang and Y. Liu, Angew. Chem., Int. Ed., 2021, 60, 11335–11343.
41. D. Wang, J. Zhang, Z. Huang, Y. Yang, T. Fu, Y. Yang, Y. Lyu, J. Jiang, L. Qiu, Z. Cao, X. Zhang, Q. You, Y. Lin, Z. Zhao and W. Tan, ACS Cent. Sci., 2023, 9, 72–83.

---

Footnote

† D. Li and Q. Li contributed equally.

---